• adults
• clinical
• disorders
• healthy
• melatonin
• meta-analysis
• neurodevelopmental
• neuropsychiatric
• review
• sleep
• trial

Introduction 
Exogenous melatonin is commonly used to improve sleep in paediatric and adult populations. Its efficacy as a sleep aid remains debated due to conflicting results, possibly due to the variability in dosage and administration timing. A recent meta-analysis suggested that administering melatonin mimicking the endogenous circadian secretion pattern (i.e., 3 hours before the desired bedtime), as well as increasing the dose up to 4 mg, maximizes its sleep-promoting properties in healthy adults and patients affected by primary insomnia. The aim of the study was to replicate these findings using two different approaches:
- a meta-analysis on the effects of melatonin on sleep parameters in neuropsychiatric, pediatric patients;
- a clinical trial in healthy adults.
Materials and Methods  
A systematic search for the meta-analysis was conducted across multiple databases (PubMed, Cochrane, Web of Science, Scopus) following PRISMA guidelines. The study outcomes were the mean differences (MD) in sleep-related variables between the treatment and placebo groups. Dose-response curves were modeled to examine the relationship between melatonin dosage and sleep variables, while meta-regression models assessed the effects of treatment duration, time between administration and bedtime, and diagnosis status. In the clinical trial, 9 volunteers underwent six overnight polysomnography evaluations under different experimental conditions:
- adaptation night
- melatonin 3 hours before bedtime, 2 mg
- melatonin 30 minutes before bedtime, 2 mg
- melatonin 3 hours before bedtime, 4 mg
- melatonin 30 minutes before bedtime, 4 mg
- placebo
Results 
The meta-analysis included 21 studies and found that melatonin significantly reduced sleep onset latency (MD = -22.76 min, p < 0.001), increased sleep efficiency (MD = 5.52%, p < 0.001) and total sleep time (MD = 38.65 min, p < 0.001), and nearly reduced wake after sleep onset (MD = -14.10 min, p = 0.06). Patients with neurodevelopmental disorders showed a greater reduction in sleep onset latency (β = -11.16, p = 0.0044). Sleep onset latency was associated with the time between melatonin intake and bedtime (β = -5.51 min, p = 0.042), while sleep efficiency and total sleep time were related to treatment duration (β = 0.78 %/week of treatment, p < 0.001; β = 5.25 min/week of treatment, p < 0.001). Melatonin dose between 2-5 mg had the greatest effect. In the clinical trial, administering 4 mg of melatonin 30 minutes before bedtime nearly significantly reduced sleep onset latency (p = 0.06) and reduced self-report WASO (p = 0.034). Administering 4 mg, 3 hours before bedtime was associated with increased REM latency (p = 0.043) and N3 sleep phase duration (p = 0.023).
Conclusions 
In pediatric neuropsychiatric patients, the administration of exo-MEL at 2-5 mg/day reduces sleep onset latency and increases sleep duration and efficiency, with earlier administration enhancing the first effect. Longer treatment further improves both sleep efficiency and total sleep time. These findings suggest potential benefits for neuropsychiatric paediatric patients, though further clinical trials are needed. In healthy adults, a 4 mg dose of melatonin effectively reduces sleep onset latency and WASO when taken 30 minutes before bed. When taken 3 hours before the desired bedtime, it also seems to increase the duration of the first sleep cycle, with potential benefits associated with boosting slow wave activity. Overall, the dose seems more influential than timing.